perm filename INTU.SQ[DIS,DBL] blob
sn#208280 filedate 1976-04-01 generic text, type C, neo UTF8
COMMENT ⊗ VALID 00002 PAGES
C REC PAGE DESCRIPTION
C00001 00001
C00002 00002 . SSSEC(The Intuition for Squaring)
C00006 ENDMK
C⊗;
. SSSEC(The Intuition for Squaring)
The notion of "Intuition" is so strange that I shall take every
opportunity to try to explicate it. This subsection is a digression
to that purpose, and may therefore be skipped forever, returned to
later, or read right now.
Consider what happens when the task ⊗6"Fill-in intuitions of
TIMES-Itself"⊗* gets chosen. Rippling upward, we find that the
intuition for TIMES is a rectangle (in the plane) built out of unit
squares, where the length of each side of the rectangle represents
the two arguments to TIMES, and the area of the rectangle (i.e., the
rectangle itself) represents the value of TIMES. For the current
operation, this translates to a rectange whose sides are all of equal
length: i.e., a square. If AM is given this notion explicitly, it
will actually suggest renaming the TIMES-Itself operation "Square".
Notice the nice intuitive image that this will cause for
square-rooting, since a number n is represented as a pile of n 1x1
squares. To take the square-root of a number, we must arrange that
pile into one large square, and then see how long one side is.
This also suggests several trivial relationships to verify. Since the
number 1 is intuited as a 1x1 square, it seems intuitively clear to
AM that Times-Itself(1) will be equal to 1. This is then checked
"officially" (using the definition of Times-Itself) and verified.
Similar considerations suggest that Times-Itself(0) will equal 0, and
that in general Times-Itself(x) is much larger than x.
Although AM never noticed it, this intuition is just the right one to
notice that to get from n⊗A2⊗* to n+1⊗A2⊗*, it is necessary to add
2n+1 unit squares: i.e., that n⊗A2⊗* + 2n + 1 = n+1⊗A2⊗*. This is a
far from trivial result to notice, when one doesn't know about
algebra. In fact, AM couldn't verify such a general conjecture, since
it knows neither the concept of mathematical induction, nor anything
about proof in general.